JP2007151041A - Interconnection switching hub - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an interconnection switching hub which rationalizes path switching. <P>SOLUTION: An active machine switched from a standby state to an active state on receipt of notification of failure from an active machine immediately broadcast-transmits an echo request frame into its own network and learns an address within the network by a returned echo response frame. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to a switching hub for connecting two networks with a redundant path, and more particularly to an interconnection switching hub that rationalizes path switching.

  In order to provide a redundant path for interconnecting two networks, two switching hubs in each network may be interconnected. If the paths are made redundant in this way, communication between networks can be prevented from being interrupted by switching to another healthy path when a failure occurs in the used path.

  As shown in FIG. 7, a path is established by connecting the two switching hubs L1 # 1 and L1 # 2 of one network L1 to the switching hubs L2 # 1 and L2 # 2 of the partner network L2 on a one-to-one basis. Redundant. In each of the networks L1 and L2, there are switching hubs L1 # 3 and L1 # 4 and a plurality of switching hubs other than those shown in the figure, and these switching hubs are connected in a network. Which switching hub is used for network interconnection is arbitrary, but here, switching hubs # 1 and # 2 used for network interconnection are referred to as interconnection switching hubs for distinction. For the sake of simplicity, it is assumed that two interconnection switching hubs # 1 and # 2 in the same network are directly connected to each other.

  Two interconnection switching hubs # 1 and # 2 belonging to the same network are divided into an active state and a standby state with respect to the inter-network relay. The switching hub for interconnection that has become active is called an active machine, and the switching hub for interconnection that has become standby is called a standby machine. As a matter of course, each set of interconnection switching hubs connected to each other between the two networks becomes an active machine. In the illustrated example, the interconnection switching hub L1 # 1 and the interconnection switching hub L2 # 1 are active machines. Thus, communication between networks is relayed using a path between active machines.

  The switching hub has a function of detecting that a failure has occurred in a transmission path connected to the switching hub. Therefore, when the active machine detects a failure that has occurred on the path between the active machines, it notifies the standby machine of the occurrence of the failure, and the standby machine switches to active according to the notification and relays communication between networks. It is like that. The original active machine switches to standby. Similarly, the standby machine switches to active in the partner network, so communication between networks is not interrupted.

JP 2000-13430 A

  In FIG. 7, it is assumed that a failure has occurred in the path between the interconnecting switching hubs L1 # 1 and L2 # 1, as shown in FIG. In other words, the interconnection switching hub L1 # 2 and the interconnection switching hub L2 # 2 that are standby machines are switched to active, and a communication path between networks is secured.

  However, the switching hub has an address learning function. The address learning function is a function of storing a transmission source MAC address written in a frame in a table in association with the port number when the frame is received at the port. The learned address is used for filtering. Filtering means that when a frame is received, the frame is relayed only from the port specified by searching the table with the destination MAC address written in the frame. When there is no corresponding address in the table, flooding is performed. Flooding refers to relaying a frame from all ports other than the port that received the frame.

  Now, if there are switching hubs L1 # 3 and L1 # 4 that are directly connected to the interconnection switching hub L1 # 1, these switching hubs L1 # 3 and L1 # 4 are transmitted from within the network L2. Since all the received frames are received from the interconnection switching hub L1 # 1, the source MAC addresses of these frames are all learned in correspondence with the ports connected to the interconnection switching hub L1 # 1. Therefore, the frame addressed in the network L2 received at the other port is relayed to the interconnection switching hub L1 # 1 by filtering.

  In this state, the switching of the network interconnection path and the relay operation by filtering are not meshed. That is, the frame addressed to the network L2 is relayed to the interconnection switching hub L1 # 1 that has already been switched to standby (in the figure, the frame addressed from the network L2 to the network L1 is switched to standby. The image gathered at the switching hub L2 # 1 is indicated by an arrow b). In order for these frames to reach the new standby machine, the switching hubs # 1 and # 2 for interconnection and all the switching hubs directly connected to the switching hubs # 1 and # 2 for interconnection must The address information obtained from past address learning must be cleared. For this reason, the switching hub has a function of notifying other switching hubs and instructing clearing of address information when it detects that a failure has occurred in a transmission line connected to the switching hub.

  As described above, there are a number of technologies such as the spanning tree protocol in which the switching hub informs each other of the occurrence of a failure and triggers clearing of the address information, and each manufacturer uses a unique method. However, since each manufacturer uses a unique method, all the switching hubs directly connected to the interconnection switching hubs # 1 and # 2 and the interconnection switching hubs # 1 and # 2 are the same manufacturer. Otherwise, the address information will not be cleared successfully. Therefore, the six switching hubs L1 # 1 to # 4, L2 # 1, and L2 # 2 written in blocks in FIG. 7 are forced to be the same manufacturer.

  As mentioned above, which switching hub is used for network interconnection is optional, but in practice it can be chosen because of the need to unify the address information clearing method (or failure occurrence notification method). Switching hubs are limited. In other words, compatibility between manufacturers is poor.

  After the address information is cleared, address learning starts again, and address learning progresses as frames are exchanged within each network and between networks. During that time, flooding is frequently performed. Become. If address learning can be achieved quickly without clearing the address information, frequent flooding can be avoided.

  SUMMARY OF THE INVENTION Accordingly, an object of the present invention is to provide an interconnection switching hub that solves the above-described problems and streamlines path switching.

  In order to achieve the above object, the present invention provides two switching hubs in each network in order to redundantly connect a partner network having the same configuration as a network formed by connecting a plurality of switching hubs in a network. Is used as a switching hub for interconnection, and the first interconnection switching hubs of both networks are interconnected one-to-one and the second interconnection switching hubs of both networks are interconnected one-to-one. In both cases, the switching hub for the first interconnection is the active machine and the switching hub for the second interconnection is the standby machine, so that the active machine relays communication between networks using the path between the active machines, and the active machine is active. When a failure in the route between aircrafts is detected, the active aircraft The failure notification is sent to the standby device, and the standby device is switched to active by the notification and relays communication between networks, and the switched active device immediately broadcasts an echo request frame in its own network. The address in the network is learned by the echo response frame transmitted and returned.

  The switched active device may relay the echo response frame returned from its own network to the partner network so that the switched active device of the partner network learns the address.

  The present invention exhibits the following excellent effects.

  (1) Route switching can be streamlined compared to the prior art.

  Hereinafter, an embodiment of the present invention will be described in detail with reference to the accompanying drawings.

  As shown in FIG. 1, each of the interconnection switching hubs according to the present invention has a network L1 in which a plurality of switching hubs are connected in a network and a partner network L2 having the same configuration as each other. Two switching hubs $ 1 and $ 2 in each of the networks L1 and L2 are used as interconnection switching hubs, and the first interconnection switching hubs $ 1 in both networks L1 and L2 are connected in a one-to-one relationship. At the same time, the switching hubs $ 2 for the second interconnections of both networks L1 and L2 are connected to each other in a one-to-one relationship. By making switching hub $ 2 a standby machine, the active machine uses the path between the active machines to When the active machine detects a failure that occurred in the path between the active machines, the failure notification is sent from the active machine to the standby machine. In addition to relaying communications, the switched active machine immediately broadcasts an echo request frame in its own network, and learns the address in the network from the returned echo response frame. .

  Further, here, the switched active device relays the echo response frame sent back from within its own network to the partner network, so that the partner's network switched active device learns the address.

  The first and second words are added to clarify the distinction in the claims, but will be omitted below because they can be distinguished by the symbols $ 1 and $ 2. In addition, a switching hub whose code is $ indicates that it is a switching hub equipped with the function of the present invention. The switching hub with the symbol # or a switching hub other than that shown need not have the function of the present invention.

  For the sake of simplicity, it is assumed that two interconnection switching hubs $ 1 and $ 2 in the same network are directly connected to each other. However, even if another switching hub exists between the interconnection switching hubs $ 1, $ 2, there is no particular influence on the present invention.

  The figure shows some terminals 101 connected to the switching hubs L1 # 3 and L1 # 4. In addition, there are terminals connected to the switching hubs $ 1 and $ 2 for interconnection and the switching hubs not shown. Needless to say, they are connected. The address in the network is the address of these switching hubs and terminals.

  The operation will be described below.

  Since the interconnection switching hubs L1 $ 1 and L2 $ 1 are active machines, the communication path between the networks L1 and L2 is between the interconnection switching hubs L1 $ 1 and L2 $ 1. Therefore, as a result of address learning in each switching hub, the interconnection switching hub L1 $ 1 stores the address in the network L2 in association with the port to which the interconnection switching hub L2 $ 1 is connected, and the switching hub. In L1 # 3, L1 # 4 and the interconnection switching hub L1 $ 2, the address in the network L2 is stored in association with the port to which the interconnection switching hub L1 $ 1 is connected.

  Here, as shown by a cross in FIG. 1, when a failure occurs in the transmission path connecting the interconnection switching hubs L1 $ 1 and L2 $ 1, the interconnection switching hubs L1 $ 1 and L2 $ 1 are When a failure occurrence is detected by a conventional failure occurrence detection function, a failure occurrence notification is given to the switching hubs L1 $ 2 and L2 $ 2 for interconnection by using a conventional failure notification function. Switch to standby. On the other hand, the interconnecting switching hubs L1 $ 2 and L2 $ 2 that have received the notification are switched to active.

  Hereinafter, since the operation of the present invention performed in the network L1 and the network L2 is the same, only the network L1 will be described.

  As shown in FIG. 2, the interconnection switching hub L1 $ 2 that has become a new active device broadcasts echo request frames from all the ports connected to the network L1. This echo request frame is received and relayed to all the switching hubs in the network L1 as well as the switching hubs L1 # 3 and L1 # 4. When each switching hub receives an echo request frame at a certain port, the switching hub broadcasts an echo response frame from all the ports. In addition, when each terminal receives the echo request frame, each terminal transmits an echo response frame by broadcast.

  FIG. 3 shows an echo frame format. The echo frame includes a destination MAC address column, a source MAC address column, an ether type column, an IP header column, and an ICMP message column. The ICMP message column includes a type column, a code column, and a checksum column. Column, identifier column, sequence number column, and data column. In the case of an echo request frame, the numerical value 8 is stored in the type column, and in the case of an echo response frame, the numerical value 0 is stored in the type column. In either case, the code column stores the numerical value 0. Since this format is defined in ICMP (Internet Control Message Protocol; RFC 792), the description of further known matters is omitted.

  The IP header column includes a destination IP address column (not shown). The main point of the present invention is to store the broadcast address in this destination IP address column. That is, when an echo request frame is transmitted, the destination MAC address and the destination IP address are broadcast addresses, and the source MAC address is a unique address that does not exist all over the world. A unique address that does not exist in the world is a MAC address that is reserved and used for the present invention by a MAC address that is not attached to a terminal or switching hub by any manufacturer in the world. Since this MAC address is not actually attached to a terminal or a switching hub, it is a unique address that does not exist in the world. As the source IP address, a unique address that does not exist in the network to which the device belongs is used.

  The switching hub and the terminal that have received such an echo request frame use a unique address that does not exist in the world, which is the source MAC address stored in the echo request frame, as the destination MAC address, and the source MAC address as the source. An echo response frame with the MAC address is transmitted.

  Since each switching hub that has received such an echo response frame is an address for which there is no destination MAC address, this echo response frame is relayed by broadcast.

  In general, broadcast relay means that the address of the destination terminal is unlearned, so even if the destination address is written in unicast, relaying from multiple ports results in broadcast. However, in this case, since the destination address is “an address that does not exist”, it means that it becomes a broadcast.

  As a result of the broadcast of the echo request frame as shown in FIG. 2, echo response frames are returned from all the switching hubs and terminals in the network L2 to the interconnection switching hub L1 $ 2. Therefore, the interconnection switching hub L1 $ 2 can learn the correspondence between the plurality of ports of the interconnection hub L1 $ 2 and all addresses in the network L2.

  In addition, the echo request frame is broadcast and relayed from the switching hub to the switching hub, and the address response learning is also performed in each switching hub in the process in which the echo response frame is relayed from the switching hub to the switching hub through the reverse route. Will be achieved.

  Further, the interconnection switching hub L1 $ 2 transmits the received echo response frame to the network L2. Since the destination MAC address of this echo response frame does not exist as described above, each switching hub in the network L2 relays this echo response frame by broadcast. The image is represented by an arrow a. Therefore, address learning also proceeds in the network L2.

  Conventionally, when the occurrence of a failure is notified, all the address information learned before that is erased, and then a general frame exchanged thereafter is flooded to relearn the address. Since the present invention actively promotes relearning by issuing an echo request frame, address learning can be achieved quickly.

  In the past, two interconnecting switching hubs and all the switching hubs directly connected to them had to be the same manufacturer. However, in the present invention, only two interconnecting switching hubs are the present invention. The other switching hubs may be of any manufacturer as long as they are equipped with ICMP, which has been generally used. That is, compatibility between manufacturers is improved.

  Next, another embodiment of the present invention will be described.

  As shown in FIG. 4, switching hubs $ 1 and $ 2 for interconnection are included in both networks L1 and L2. More specifically, some of the plurality of ports of the interconnection switching hub $ 1 are connected to the switching hub or terminal in the network L1, and the remaining some ports are connected to the switching hub or terminal in the network L2. It is connected.

  Even in such a form, the switching hubs for interconnection $ 1 and $ 2 operate in the same manner as described above. That is, when the interconnection switching hub $ 1 that was an active device detects a failure in the transmission path at either port of the network L1 or L2, the interconnection switching hub $ 2 is switched to the active state by the notification. Interconnection switching hub $ 2 broadcasts an echo request frame in both networks L1 and L2, which are its own networks. Therefore, address relearning is promoted in each of the networks L1 and L2.

  In this embodiment as well, only the switching hubs $ 1 and $ 2 for interconnection are sufficient to mount the function of the present invention, and the switching hubs L1 # 3 and L1 # 4 and other switching hubs (not shown) are manufacturers and models. Do not choose.

  As shown in FIG. 5, when there are three networks L1, L2, and L3, the interconnection switching hubs $ 1 and $ 2 of the present invention are provided in each of the networks L1, L2, and L3. Between the networks L1 and L2, the interconnection switching hubs $ 1 are connected to each other, and the interconnection switching hubs $ 2 are connected to each other. Also, the interconnection switching hubs $ 1 are connected to each other between the networks L1 and L3, and the interconnection switching hubs $ 2 are connected to each other.

  The operation in the networks L1 and L2 when a failure occurs between the networks L1 and L2 is as described above. Since the operation in the networks L1 and L3 when a failure occurs between the networks L1 and L3 is the same, the description will not be repeated.

  When a failure occurs between the networks L1 and L2, if the failure is caused by a failure of a device in the network L1, a failure also occurs between the networks L1 and L3. Therefore, the switching hubs L1 $ 1 and L3 for interconnection $ 1 switches to standby.

  When a failure occurs between the networks L1 and L2, if the failure is caused by a failure in the transmission path between the interconnection switching hubs L1 $ 1 and L2 $ 1, the failure does not spread between the networks L1 and L3. At this time, the interconnection switching hub L3 $ 1 is switched to the standby state by the operation of a protocol different from the present invention.

  As described above, the present invention is effective even when the paths interconnecting the three networks L1, L2, and L3 are made redundant. Needless to say, the present invention is effective even when the number of networks is four or more.

  Next, the internal structure of the interconnection switching hub of the present invention will be described.

  As shown in FIG. 6, the switching hub 61 according to the present invention includes a plurality of ports 62, a PHY (physical interface) 63 provided for each port, a communication control circuit (MAC) 64, and a learned address. It includes an FDB 65 that stores information, and a CPU 66 that processes processing related to the present invention and other logical protocols. Some of the ports are connected to their own network (indicated as an internal network in the figure), and some are connected to a partner network (indicated as a separate network in the figure).

  The CPU 66 generates an echo request frame and outputs it to a port connected to its own network in the operation flow described so far. The PHY 63 determines whether or not the transmission path is normal from the current and voltage of the transmission path. From this determination result, the CPU 66 recognizes the occurrence of a failure.

It is a network block diagram which made the interconnection path | route of two networks redundant using the switching hub for interconnection of this invention. FIG. 2 is a configuration diagram of a network after a failure occurs in the configuration of FIG. 1. It is a format diagram of an echo frame. It is the network block diagram which showed other embodiment of this invention. It is the network block diagram which showed other embodiment of this invention. It is an internal block diagram of the switching hub of this invention. It is the network block diagram which made the interconnection path | route of two networks redundant using the conventional switching hub. FIG. 8 is a configuration diagram of a network after a failure occurs in the configuration of FIG.

Explanation of symbols

101 terminal L1, L2 network L1 # 3, L1 # 4 switching hub $ 1, $ 2 (L1 $ 1, L1 $ 2, L2 $ 1, L2 $ 2) switching hub for interconnection

Claims (2)

  In order to make a redundant connection to a partner network having the same configuration as a network formed by connecting a plurality of switching hubs in a network form, two switching hubs in each network are used as switching hubs for interconnection. The first interconnection switching hubs are interconnected in a one-to-one relationship, the second interconnection switching hubs in both networks are interconnected in a one-to-one relationship, and the first interconnection switching hub is active in both networks. By using the second interconnection switching hub as a standby machine, the active machine relays communication between networks using the path between the active machines, and the active machine detects a failure that occurred in the path between the active machines. Occasionally, a failure notification is sent from the active machine to the standby machine. In response to this notification, the standby machine is actively switched and relays communication between networks, and the switched active machine immediately broadcasts an echo request frame in its own network, and the echo response frame returned. An interconnection switching hub characterized by learning an address in a network.   The switched active device relays an echo response frame returned from within its own network to the partner network so that the switched active device of the partner network learns the address. The switching hub for interconnection according to Item 1.

Images